P-E

v1.0.0

极速图片提取工具。发送模板图片+多张实际图片,秒懂格式,一步到位生成 Excel 表格。无需确认,直接输出。 触发方式:/p-e、/pe、「图片提取」「提取Excel」「picture extract」 Picture Extract: send template + data images, instantly...

0· 113·0 current·0 all-time

Install

OpenClaw Prompt Flow

Install with OpenClaw

Best for remote or guided setup. Copy the exact prompt, then paste it into OpenClaw for haiou-max/p-e.

Previewing Install & Setup.
Prompt PreviewInstall & Setup
Install the skill "P-E" (haiou-max/p-e) from ClawHub.
Skill page: https://clawhub.ai/haiou-max/p-e
Keep the work scoped to this skill only.
After install, inspect the skill metadata and help me finish setup.
Required binaries: python3
Use only the metadata you can verify from ClawHub; do not invent missing requirements.
Ask before making any broader environment changes.

Command Line

CLI Commands

Use the direct CLI path if you want to install manually and keep every step visible.

OpenClaw CLI

Bare skill slug

openclaw skills install p-e

ClawHub CLI

Package manager switcher

npx clawhub@latest install p-e
Security Scan
VirusTotalVirusTotal
Benign
View report →
OpenClawOpenClaw
Benign
high confidence
Purpose & Capability
The skill's name/description (image→Excel) matches the provided code and runtime instructions. It only requires python3 and includes a generate_excel.py that embeds images into an XLSX file — a reasonable and proportionate requirement for the stated function.
Instruction Scope
Instructions are specific: save images to /tmp/p-e-images, write JSON to /tmp/p-e-data.json, run the bundled generator. The extraction step relies on the agent's vision capability (no OCR library in the package), which is expected for an agent skill but is operationally vague. The skill intentionally retains uploaded images in /tmp rather than removing them; this may leave sensitive images on disk if the environment does not auto-clean /tmp.
Install Mechanism
No remote install spec in the registry; the bundled setup.sh uses pip3 --user to install openpyxl and Pillow from PyPI. This is a standard, low-risk approach (no downloads from unknown hosts or extract actions).
Credentials
The skill declares no environment variables, credentials, or config paths. Runtime instructions use only standard filesystem paths (/tmp and user's Desktop) and the expected python3 binary — there are no unrelated secrets requested.
Persistence & Privilege
The skill is not flagged always:true and does not attempt to modify other skills or system-wide configs. It writes files under /tmp and (by default) the user's Desktop, which is appropriate for its function but means artifacts persist until removed.
Assessment
This skill appears to do what it says: convert structured data from images into an Excel file using the included Python script. Before installing or running it, consider: 1) It will save uploaded images to /tmp/p-e-images and leave them there — avoid sending highly sensitive images or remove the files after use. 2) The default output is ~/Desktop/product_list.xlsx and may overwrite an existing file with that name; specify an output path if needed. 3) The setup step runs pip3 --user install openpyxl Pillow (PyPI packages) — if you prefer, run scripts/setup.sh yourself in a controlled environment. 4) The actual data extraction (parsing fields from images) is performed by the agent's vision/logic (not an included OCR library), so results depend on the agent's capabilities; test on non-sensitive samples first. If any of these behaviors are unacceptable (retaining files, installing packages, or running commands), do not install or run the skill until you can run it in an isolated environment or adjust the workflow.

Like a lobster shell, security has layers — review code before you run it.

Runtime requirements

📊 Clawdis
OSmacOS · Linux
Binspython3
latestvk977dhdr4g5ephcwwjrc8tvrpd84gprg
113downloads
0stars
1versions
Updated 2w ago
v1.0.0
MIT-0
macOS, Linux

P-E: 图片批量提取 → Excel 表格

概述

从相同格式的图片中提取数据,自动生成带嵌入图片的 Excel 表格。零交互,一步完成。

触发条件

当用户发送以下内容时触发:

  • 模板图片 + 多张实际数据图片
  • 说「图片提取」「提取Excel」「picture extract」
  • 使用 /p-e/pe

工作流程

第 0 步:依赖检查(仅首次)

exec 工具运行:

python3 -c "import openpyxl; import PIL" 2>/dev/null || pip3 install --user openpyxl Pillow

如果失败,告知用户运行:bash ~/.agents/skills/p-e/scripts/setup.sh

第 1 步:分析模板图片

用你的视觉能力分析用户发送的第一张(模板)图片:

  1. 识别所有字段的位置和格式
  2. 理解数据的排版规律
  3. 推断字段名称和数据类型
  4. 建立字段-位置的映射关系

第 2 步:提取所有图片数据

逐张分析实际数据图片,提取每张图片中的结构化数据。

同时,将用户发送的每张图片保存到临时目录,用于后续嵌入 Excel:

# 用 exec 工具创建临时目录
mkdir -p /tmp/p-e-images

如果图片已经在本地文件系统上(用户给了路径),直接使用该路径。 如果图片是通过聊天发送的,用 write 工具将图片保存到 /tmp/p-e-images/

第 3 步:生成 JSON 数据文件

将提取的数据组织为 JSON 格式,用 write 工具写入临时文件:

{
  "fields": ["编号", "产品描述", "装箱数", "单价", "尺寸", "图片"],
  "data": [
    {
      "编号": "3-4",
      "产品描述": "手掌最后的晚餐",
      "装箱数": 24,
      "单价": 25,
      "尺寸": "17×24",
      "图片": "/tmp/p-e-images/img1.jpg"
    }
  ]
}

重要规则:

  • 最后一个字段必须命名为「图片」「image」或「图」
  • 图片字段的值必须是图片文件的绝对路径
  • 数字类型(价格、数量)使用数字而非字符串
  • 无法识别的字段标记为 [待确认]

write 工具将 JSON 写入 /tmp/p-e-data.json

第 4 步:运行 Excel 生成脚本

exec 工具运行:

python3 ~/.agents/skills/p-e/scripts/generate_excel.py \
  --json /tmp/p-e-data.json \
  --output ~/Desktop/product_list.xlsx

如果用户指定了输出路径,使用用户指定的路径。默认输出到桌面。

第 5 步:清理临时文件

rm -f /tmp/p-e-data.json

保留 /tmp/p-e-images/ 中的图片(Excel 生成后不再需要,但用户可能想保留)。

第 6 步:报告结果

输出格式:

✓ 已识别 N 张图片

✓ 已生成 product_list.xlsx

数据统计:
- N 行数据
- M 列字段(字段1|字段2|...|图片)
- 所有图片自动嵌入到最后一列
- 完毕!

Excel 输出规格

属性
文件名product_list.xlsx(默认)
表头样式蓝色背景 (#4472C4),白色粗体字
数据列宽15
图片列宽55
行高190
图片尺寸320×240px,原图直接嵌入不压缩
边框所有单元格细边框

异常处理

情况处理方式
字段不清晰或缺失标记为 [待确认]
一张图有多条数据自动拆分为多行
信息缺失该单元格留空
图片文件不存在显示 [File not found]
Python 依赖缺失自动安装或提示运行 setup.sh

使用示例

用户发送:

【模板】
[鞋子信息表照片]

【实际数据】
[鞋1] [鞋2] [鞋3] [鞋4]

或者指定文件夹:

/path/to/images 帮我把这些图片整理出表格

Comments

Loading comments...